U.S. patent application number 14/864300 was filed with the patent office on 2016-12-01 for methods and apparatus to assign demographic information to panelists.
The applicant listed for this patent is The Nielsen Company (US), LLC. Invention is credited to Michelle Christian, Shelli Kashriel, ChoongKoo Lee, Jonathan Sullivan.
Application Number | 20160350556 14/864300 |
Document ID | / |
Family ID | 57398872 |
Filed Date | 2016-12-01 |
United States Patent
Application |
20160350556 |
Kind Code |
A1 |
Sullivan; Jonathan ; et
al. |
December 1, 2016 |
METHODS AND APPARATUS TO ASSIGN DEMOGRAPHIC INFORMATION TO
PANELISTS
Abstract
Methods and apparatus to assign demographic information to
panelists are disclosed. Example disclosed methods include
generating decoy database proprietor identifiers based on
probability density functions. The example method also include
querying a database proprietor using panelist database proprietor
identifiers and the decoy database proprietor identifiers to obtain
demographic information associated with the panelist database
proprietor identifiers. The example method also include assigning
the panelist database proprietor identifiers to panelists based on
the demographic information obtained from the database
proprietor.
Inventors: |
Sullivan; Jonathan; (Natick,
MA) ; Kashriel; Shelli; (New York, NY) ;
Christian; Michelle; (Oldsmar, FL) ; Lee;
ChoongKoo; (Schaumburg, IL) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
The Nielsen Company (US), LLC |
New York |
NY |
US |
|
|
Family ID: |
57398872 |
Appl. No.: |
14/864300 |
Filed: |
September 24, 2015 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
62167820 |
May 28, 2015 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G06F 21/6254 20130101;
G06Q 30/0204 20130101; G06F 16/2457 20190101 |
International
Class: |
G06F 21/62 20060101
G06F021/62; G06F 17/30 20060101 G06F017/30; G06Q 30/02 20060101
G06Q030/02 |
Claims
1. A method comprising: generating decoy database proprietor
identifiers based on probability density functions; querying a
database proprietor using panelist database proprietor identifiers
and the decoy database proprietor identifiers to obtain demographic
information associated with the panelist database proprietor
identifiers; and assigning the panelist database proprietor
identifiers to panelists based on the demographic information
obtained from the database proprietor.
2. A method as defined in claim 1, further including randomly
arranging the panelist database proprietor identifiers and the
decoy database proprietor identifiers into a query to query the
database proprietor.
3. A method as defined in claim 1, further including determining
the probability density functions based on independent
distributions of numbers within the panelist database proprietor
identifier.
4. A method as defined in claim 1, including determining the
probability density functions based on dependent distributions of
numbers within the panelist database proprietor identifier.
5. A method as defined in claim 1, wherein the generating of the
decoy database proprietor identifiers based on the probability
density functions further includes, for each of digits of a
corresponding one of the decoy database proprietor identifiers,
selecting a numeric value for the digit based on a corresponding
one of the probability density functions.
6. A method as defined in claim 1, wherein a quantity of the decoy
database proprietor identifiers are generated based on a confidence
level of the probability density functions.
7. A method as defined in claim 1, wherein, in response to querying
the database proprietor using the panelist database proprietor
identifiers and the decoy database proprietor identifiers:
receiving (a) demographic information corresponding to the panelist
database proprietor identifiers and the decoy database proprietor
identifiers that are associated with subscribers to the database
proprietor, and (b) errors indicating that particular ones of the
decoy database proprietor identifiers are not associated with the
subscribers to the database proprietor; and regenerating the
probability density functions based on the panelist database
proprietor identifiers and the decoy database proprietor
identifiers that are associated with the subscribers to the
database proprietor.
8. An apparatus comprising: a decoy generator to generate decoy
database proprietor identifiers based on probability density
functions; a query handler to query a database proprietor using
panelist database proprietor identifiers and the decoy database
proprietor identifiers to obtain demographic information associated
with the panelist database proprietor identifiers; and a panelist
comparator to assign the panelist database proprietor identifiers
to panelists based on the demographic information obtained from the
database proprietor.
9. An apparatus as defined in claim 8, further including a panelist
obscurer to randomly arrange the panelist database proprietor
identifiers and the decoy database proprietor identifiers into a
query to query the database proprietor.
10. An apparatus as defined in claim 8, wherein the decoy generator
is further to determine the probability density functions based on
independent distributions of numbers within the panelist database
proprietor identifier.
11. An apparatus as defined in claim 8, including the decoy
generator is further to determine the probability density functions
based on dependent distributions of numbers within the panelist
database proprietor identifier.
12. An apparatus as defined in claim 8, wherein to generate the
decoy database proprietor identifiers based on the probability
density functions, the decoy generator is to, for each of digits of
a corresponding one of the decoy database proprietor identifiers,
selecting a numeric value for the digit based on a corresponding
one of the probability density functions.
13. An apparatus as defined in claim 8, wherein the decoy generator
is to generate a quantity of the decoy database proprietor
identifiers based on a confidence level of the probability density
functions.
14. An apparatus as defined in claim 8, wherein, in response to
querying the database proprietor using the panelist database
proprietor identifiers and the decoy database proprietor
identifiers: the query handler is to receive (a) demographic
information corresponding to the panelist database proprietor
identifiers and the decoy database proprietor identifiers that are
associated with subscribers to the database proprietor, and (b)
errors indicating that particular ones of the decoy database
proprietor identifiers are not associated with the subscribers to
the database proprietor; and the decoy generator is to regenerate
the probability density functions based on the panelist database
proprietor identifiers and the decoy database proprietor
identifiers that are associated with the subscribers to the
database proprietor.
15. A tangible computer readable storage medium comprising
instructions which, when executed, cause a machine to at least:
generate decoy database proprietor identifiers based on probability
density functions; query a database proprietor using panelist
database proprietor identifiers and the decoy database proprietor
identifiers to obtain demographic information associated with the
panelist database proprietor identifiers; and assign the panelist
database proprietor identifiers to panelists based on the
demographic information obtained from the database proprietor.
16. A tangible computer readable storage medium as defined in claim
15, wherein the instructions cause the machine to randomly arrange
the panelist database proprietor identifiers and the decoy database
proprietor identifiers into a query to query the database
proprietor.
17. A tangible computer readable storage medium as defined in claim
15, wherein the instructions cause the machine to determine the
probability density functions based on independent distributions of
numbers within the panelist database proprietor identifier.
18. A tangible computer readable storage medium defined in claim
15, wherein the instructions cause the machine to determine the
probability density functions based on dependent distributions of
numbers within the panelist database proprietor identifier.
19. A tangible computer readable storage medium as defined in claim
15, wherein to generate the decoy database proprietor identifiers
based on the probability density functions, the instructions cause
the machine to, for each of digits of a corresponding one of the
decoy database proprietor identifiers, selecting a numeric value
for the digit based on a corresponding one of the probability
density functions.
20. A tangible computer readable storage medium as defined in claim
15, wherein the instructions cause the machine to generate a
quantity of the decoy database proprietor identifiers based on a
confidence level of the probability density functions.
21. A tangible computer readable storage medium as defined in claim
15, wherein, in response to receiving the demographic information
associated with subscribers to the database proprietor
corresponding to the decoy database proprietor identifiers, the
instructions cause the machine to regenerate the probability
density functions based on the panelist database proprietor
identifiers and the decoy database proprietor identifiers that are
associated with the subscribers to the database proprietor.
Description
RELATED APPLICATIONS
[0001] This patent claims benefit to U.S. Provisional Application
Ser. No. 62/167,820, filed May 28, 2015, which is herein
incorporated by reference in its entirety.
FIELD OF THE DISCLOSURE
[0002] This disclosure relates generally to audience measurement
and, more particularly, to methods and apparatus to assign
demographic information to panelists.
BACKGROUND
[0003] Audience measurement entities measure exposure of audiences
to media such as television, music, movies, radio, Internet
websites, streaming media, etc. The audience measurement entities
generate ratings based on the measured exposure. Ratings are used
by advertisers and/or marketers to purchase advertising space
and/or design advertising campaigns. Additionally, media producers
and/or distributors use the ratings to determine how to set prices
for advertising space and/or to make programming decisions.
[0004] Techniques for monitoring user access media have evolved
significantly over the years. Some prior systems perform such
monitoring primarily through server logs. In particular, entities
serving media on the Internet can use such prior systems to log the
number of requests received for their media at their server.
BRIEF DESCRIPTION OF THE DRAWINGS
[0005] FIG. 1 illustrates an example system to assign demographic
information to panelists.
[0006] FIG. 2 illustrates the example demographic retriever of FIG.
1 to retrieve panelist demographic information from the example
database proprietor of FIG. 1.
[0007] FIG. 3 illustrates example database proprietor identifiers
and example decoy database proprietor identifiers used by the
example demographic retriever of FIG. 1 to retrieve panelist
demographic information from the example database proprietor of
FIG. 1.
[0008] FIG. 4 illustrates the example panelist comparator of FIG. 1
to assign a database proprietor identifier to a member of a
panelist household.
[0009] FIG. 5 is a flow diagram representative of example machine
readable instructions that may be executed to implement the example
demographic retriever of FIGS. 1 and 2 and/or the example panelist
comparator of FIGS. 1 and 4 to assign demographic information to
panelists.
[0010] FIG. 6 is a flow diagram representative of example machine
readable instructions that may be executed to implement the example
decoy generator of FIG. 2 to generate probability density functions
(PDFs) used to generate decoy database proprietor identifiers.
[0011] FIG. 7 is a flow diagram representative of example machine
readable instructions that may be executed to implement the example
panelist comparator of FIGS. 1 and 4 to assign a database
proprietor identifier to a member of a panelist household.
[0012] FIG. 8 illustrates an example processor system structured to
execute the example instructions represented in FIGS. 5, 6, and/or
7 to implement the example demographic retriever of FIGS. 1 and/or
2, the example decoy generator of FIG. 2, and/or the example
panelist comparator of FIGS. 1 and/or 4.
DETAILED DESCRIPTION
[0013] Examples disclosed herein may be used to assign a database
proprietor identifier to a member of a panelist household. To
measure audiences, an audience measurement entity (AME) may use
instructions (e.g., Java, java script, or any other computer
language or script) embedded in media to collect information
indicating when audience members are accessing media on a computing
device (e.g., a computer, a laptop, a smartphone, a tablet, etc.).
Media to be traced is tagged with these instructions. When a device
requests the media, both the media and the instructions are
downloaded to the client. The instructions cause information about
the media access to be sent from the device to a monitoring entity
(e.g., the AME). Examples of tagging media and tracing media
through these instructions are disclosed in U.S. Pat. No.
6,108,637, issued Aug. 22, 2000, entitled "Content Display
Monitor," which is incorporated by reference in its entirety
herein.
[0014] Additionally, the instructions cause one or more user and/or
device identifiers (e.g., an international mobile equipment
identity (IMEI), a mobile equipment identifier (MEID), a media
access control (MAC) address, an app store identifier, an open
source unique device identifier (OpenUDID), an open device
identification number (ODIN), a login identifier, a username, an
email address, user agent data, third-party service identifiers,
web storage data, document object model (DOM) storage data, local
shared objects also referred to as "Flash cookies"), browser
cookies, an automobile vehicle identification number (VIN), etc.)
located on the computing device to be sent to a partnered database
proprietor (e.g., Facebook, Twitter, Google, Yahoo!, MSN, Apple,
Experian, etc.) to identify demographic information (e.g., age,
gender, geographic location, race, income level, education level,
religion, etc.) for the audience member of the computing device
collected via a user registration process. For example, an audience
member may be viewing an episode of "Modern Family" in a media
streaming app on a tablet device. In that instance, in response to
instructions executing within the app, a user/device identifier
stored on the tablet device is sent to the AME and/or a partner
database proprietor to associate the instance of media exposure
(e.g., an impression) to corresponding demographic information of
the audience member. The database proprietor can then send logged
demographic impression data to the AME for use by the AME in
generating, for example, media ratings and/or other audience
measures.
[0015] In some examples, the partner database proprietor does not
provide individualized demographic information (e.g., user-level
demographics) in association with logged impressions. Instead, in
some examples, the partnered database proprietor provides aggregate
demographic impression data (sometimes referred to herein as
"aggregate census data"). For example, the aggregate demographic
impression data provided by the partner database proprietor may
show that a hundred thousand females age 17-45 watched the episode
of "Modern Family" in the last seven days via computing devices
(e.g., desktop computers, tables, smart phones, laptops, etc.).
However, the aggregate demographic information from the partner
database proprietor does not identify individual persons (e.g., is
not user-level data) associated with individual impressions. In
this manner, the database proprietor protects the privacies of its
subscribers/users by not revealing their identities and, thus,
user-level media access activities, to the AME.
[0016] The AME uses this aggregate demographic information to
calculate ratings and/or other audience measures for corresponding
media. However, during the process of registering with the database
proprietor, a subscriber may lie or may otherwise provide
inaccurate demographic information. For example, during
registration, the subscriber may provide an inaccurate age or
location. These inaccuracies cause errors in the aggregate
demographic information from the partner database proprietor, and
can lead to errors in audience measurement. To combat these errors,
the AME recruits panelist households that consent to monitoring of
their exposure to media. During the recruitment process, the AME
obtains detailed demographic information from the members of the
panelist household. The AME compares the detailed demographics to
the demographic information the members of the panelist household
supplied to the database proprietor(s) to predict how demographic
information is inaccurate (e.g., misattributed) within the
aggregate demographic information provided by the database
proprietors. The example AME generates misattribution correction
factors to be applied to the aggregate demographic information
provided by the database proprietors to correct for the
inaccuracies.
[0017] In examples disclosed herein, to retrieve the demographic
information from the database proprietors, the AME obtains panelist
database proprietor identifiers (DPIDs) for members the panelist
household. As used herein, the DPID is an alphanumeric value
assigned to the database proprietor subscriber when subscriber
registers with the database proprietor. The DPID is used internally
by the database proprietor to uniquely identify the subscriber. In
some examples, the DPID is different than a subscriber chosen
identifier (e.g., a username, an authentic name, etc.). For
example, a first database proprietor subscriber named "Adam Smith"
may have an assigned DPID of "44698599407828," and a second
database proprietor subscriber named "Adam Smith" may have an
assigned DPID of "11790906116306." In some examples, the AME may,
as part of the recruitment process, ask the members of the panelist
household to provide their DPIDs. Alternatively or additionally,
the AME may retrieve the panelist DPIDs from the computing devices
of the panelist household. For example, the AME may extract the
panelist DPIDs from "cookies" deposited on the computing device
when a member of the panelist household visits the website of the
database proprietor.
[0018] In some examples, the AME obtains a list of subscribers to a
particular database proprietor and uses the panelist DPIDs to
harvest the demographic information from the list. Alternately or
additionally, the AME may use the panelist DPIDs to retrieve the
demographic information of the members of the panelist household
via an application programming interface (API) provided by the
database proprietor. That is, the AME may use the API to retrieve
the demographic information corresponding to panelists from a
subscriber database of the database proprietor. However, using the
API, the database proprietor may identify the members of the
panelist households. For example, the data base proprietor may
assume all queries coming from Internet Protocol (IP) addresses
associated with the AME are queries for members of the panelist
households.
[0019] In examples disclosed herein, to protect the privacy of
panelist households from the database proprietor, the AME generates
decoy DPIDs to mix with the panelist DPIDs. When the AME uses the
API to retrieve the demographic information from the database
proprietor, the AME creates a batch query that includes panelist
DPIDs distributed (e.g., randomly distributed, pseudo-randomly
distributed, etc.) amongst the decoy DPIDs according to an
obfuscation target (e.g. a ratio). The obfuscation target defines a
minimum number of decoy DPIDs that are to be in the batch query
based on the number of panelist DPIDs to be queried. After a batch
query is processed, the AME stores the demographic information
associated with the panelist DPIDs and discards demographic
information associated with the decoy DPIDs.
[0020] In examples disclosed herein, to generate the decoy DPIDs,
the AME creates a probability density function (PDF) for one or
more of the digit positions in the DPID. For example, if a DPID is
a fourteen-digit numeric value, the AME creates fourteen PDFs, each
corresponding to a position of one of the fourteen digits of the
DPID. The probability distributions are generated based on the
panelist household DPIDs retrieved from panelist household
computing devices. For example, the AME may determine that the
probability that the first digit position in the DPID is zero is
5.2%. To generate the decoy DPIDs, the AME samples (e.g., via
inverse transform sampling, etc.) the probability density functions
to produce a value for the corresponding digit positions.
[0021] In examples disclosed herein, after retrieving the
demographic information from the database provider and discarding
the demographic information associated with the decoy DPIDs, the
AME assigns the panelist DPID to a particular member of the
panelist household from which the panelist DPID was obtained. The
AME compares the demographic information from the database
proprietor to the demographic information of the members of the
panelist household that was collected by the AME during, for
example, the panel registration process. In some examples, the AME
compares the given names and/or variants of the given names of the
members of the panelist household to the names and/or past names
included with the demographic information from the database
proprietor. For example, if the given name of a member of a
panelist household is "James," the AME also uses one or more of
"James," "Jaime," "Jamie," "Jamey," "Jim," "Jimmy," "Jimi,"
"Jimmie," "Jay," etc. when comparing the given name of the member
of the panelist household to the names and/or the past names
included with the demographic information. In some examples, if the
given name or any of the name variants of the member of the
panelist household does not equal any of the names or the past
names included with the demographic information, the AME compares
the date of birth in the demographic information from the database
proprietor to the birth dates of members of the panelist household.
In some examples, the AME determines that the DPID is not
associated with a member of the panelist household. For example, a
friend may have logged into the database proprietor from a
computing device in the panelist household. As such, although the
friend has a DPID at the database proprietor, the friend's
demographics stored with the DPID with the database proprietor will
not match panelist demographics stored at the AME for the panelist
household.
[0022] FIG. 1 illustrates an example system 100 to assign
demographic information to panelists. In the illustrated example,
an AME 102 provides a collector 103, a DPID extractor 104, and an
AME identifier (AME ID) 106 to a computing device 108 (e.g., a
desktop, a laptop, a tablet, a smartphone, etc.) associated with a
panelist household. For example, the AME 102 may provide the
collector 103, the DPID extractor 104, and the AME ID 106 via a
registration website. In some examples, the collector 103, the DPID
extractor 104 are performed by instructions (e.g., Java, java
script, or any other computer language or script) embedded in the
registration website, or any other suitable website. In some
examples, the AME ID 106 is a cookie or is encapsulated in a cookie
set in the computing device 108 by the AME 102. Alternatively, the
AME ID 106 could be any other user and/or device identifier. In any
case, the example AME ID 106 is an alphanumeric value that the AME
102 uses to identify the panelist household.
[0023] In the illustrated example, when a member of the panelist
household uses the computing device 108 to visit a website and/or
use an app associated with a database proprietor 110, the database
proprietor 110 sets or otherwise provides, on the computing device
108, the panelist DPID 112 associated with subscriber credentials
(e.g., user name and password, etc.) used to access the website
and/or the app. In some examples, the panelist DPID 112 is a cookie
or is encapsulated in a cookie. Alternatively, the panelist DPID
112 could be any other user and/or device identifier. The example
DPID extractor 104 extracts the DPID 112 (e.g., from a cookie,
etc.). The example collector 103 collects the panelist DPIDs 112 on
the computing device 108 and sends an example ID message 114 to the
example AME 102. In the illustrated example, the ID message 114
includes the extracted panelist DPID 112 and the AME ID 106
corresponding to the panelist household. In some examples, the DPID
extractor 104 remembers the DPIDs 112 that have been extracted and
sends the ID message package 114 when a new panelist DPID 112 has
been extracted.
[0024] In the illustrated example of FIG. 1, the AME 102 stores the
extraction package 114 in a DPID database 116. The example
extraction packages 114 in the DPID database 116 are from multiple
panelist households (e.g., received from computing devices 108
associated with different panelist households, etc.). In the
illustrated example, the AME 102 includes a demographic retriever
118 structured to retrieve database proprietor demographic
information 120a from the database proprietor 110 for the panelist
DPID(s) 112 associated with panelist households. The example
demographic retriever 118 retrieves the panelist DPIDs 112 from the
example DPID database 116. The example demographic retriever 118
generates decoy DPIDs 122a, 122b used to obscure the DPIDs 112
associated with panelist households. The number of decoy DPIDs
122a, 122b generated by the example demographic retriever 118 is
based on the number of the panelist DPIDs 112 associated with
panelist households being queried and the accuracy of the decoy
DPIDs 122a, 122b generation process.
[0025] The example demographic retriever 118 randomly or
pseudo-randomly mixes the panelist DPIDs 112 and the decoy DPIDs
122a, 122b to form a batch query 123. In the illustrated example,
the database proprietor 110 provides an application program
interface (API) that provides access to the database proprietor
demographic information 120a, 120b based on DPIDs (e.g., the
panelist DPIDs 112, the decoy DPIDs 122a, 122b, etc.). To retrieve
the database proprietor demographic information 120a associated
with the panelist DPIDs 112, the example demographic retriever 118
sends the batch query 123 to the example database proprietor 110.
In response to the batch query 123, the database proprietor 110
returns a query response 125. For a particular panelist DPID 112,
the query response 124 includes the database proprietor demographic
information 120a associated with the particular panelist DPID 112.
For the decoy DPIDs 122a, 122b, the query response 124 includes
either (i) the database proprietor demographic information 120b
associated with the decoy DPID 122a (e.g., by happenstance the
decoy DPID 122a corresponds to a real subscriber), or (ii) an error
125 (e.g., the decoy DPID 122b does not correspond to a real
subscriber). When the query response 124 includes the database
proprietor demographic information 120b associated with the decoy
DPID 122a, the example demographic retriever 118 discards database
proprietor demographic information 120b associated with the decoy
DPID 122a (e.g., releases a portion of memory that is storing the
database proprietor demographic information 120b associated with
the decoy DPID 122a allowing that portion of the memory to be
overwritten).
[0026] The example AME 102 includes an example panelist comparator
126 to associate the panelist DPIDs 112 retrieved from the
computing devices 108 of panelist households to members of the
panelist households. For example, a panelist household may have
four members (e.g., a father, a father, a son, a daughter, etc.)
that are separate subscribers to the database proprietor 110. In
such an example, the DPID extractor 104 may, over time, send
multiple extraction packages 114 with each one of the extraction
packages 114 associated with the panelist DPID 112 of one of the
four members of the panelist household. The example panelist
comparator 126 compares the database proprietor demographic
information 120a associated with the panelist DPID 112 with the
demographic information of the members of the panelist household
associated with the AME ID 106 (e.g., from the corresponding
extraction package 114). In the illustrated example, demographic
information of the members of the panelist households is stored in
an example panelist database 128.
[0027] The example panelist comparator 126 compares the example
database proprietor demographic information 120a associated with
the panelist DPID 112 to the demographic information of the members
of the panelist household associated with the AME ID 106 to
determine whether the panelist DPID 112 and the database proprietor
demographic information 120a corresponds to a member of the
panelist household. If the database proprietor demographic
information 120a corresponds to the demographic information of one
of the members of the panelist household, the database proprietor
demographic information 120a and the panelist DPID 112 are stored
in the example panelist database 128 in association with the AME ID
106 of the member of the panelist household and/or an identifier of
the member of the panelist household (e.g. a panelist ID).
[0028] In the illustrated example, the AME 102 includes a
misattribution calculator 130 to estimate errors (e.g., presence of
errors and/or amounts of errors) in the database proprietor
demographic information 120a based on the differences between the
database proprietor demographic information 120a and the
corresponding demographic information of the members of the
panelist households. In the illustrated example, the demographic
information in the panelist database 128 is considered to be highly
accurate because the AME 102 collects highly accurate demographic
information from the panelist households when the members of the
panelist households consent to detailed monitoring of their access
to media on computing devices (e.g., the computing device 108). As
such, the misattribution calculator 130 considers differences
between the database proprietor demographic information 120a and
the corresponding demographic information in the panelist database
128 to be errors in the database proprietor demographic information
120a. For example, the database proprietor demographic information
120a for a member of the panelist household may indicate that the
member is thirteen when the corresponding demographic information
in the panelist database 128 indicates that the member is ten. In
the illustrated example, the misattribution calculator 130 analyzes
the database proprietor demographic information 120a relative to
the demographic information in the panelist database 128 in the
aggregate. For example, misattribution calculator 130 may detect
that 1.2% of the ages of males with reported ages (e.g., ages
reported to the database proprietor 110) of 13-16 are inaccurate by
one year, 0.7% of the ages of males with reported ages of 13-16 are
inaccurate by two years, etc.
[0029] The example misattribution calculator 130 generates
misattribution correction factors used to correct the aggregate
exposure data provided by the database proprietor 110. Examples
disclosed herein may be used in connection with techniques for
generating misattribution correction factors are disclosed in U.S.
patent application Ser. No. 14/560,947, filed Dec. 4, 2014,
entitled "Methods and Apparatus to Compensate Impression Data for
Misattribution and/or Non-Coverage by a Database Proprietor," U.S.
patent application Ser. No. 14/569,474, filed Dec. 12, 2014,
entitled "Method and Apparatus to Generate Electronic Mobile
Measurement Census Data," and U.S. patent application Ser. No.
14/604,394, filed Jan. 23, 2015, entitled "Methods and Apparatus to
Correct Age Misattribution in Media Impressions," which are
incorporated by reference in their entirety herein.
[0030] FIG. 2 illustrates the example demographic retriever 118 of
FIG. 1 to retrieve example database proprietor demographic
information 120a, 120b from the example database proprietor 110 of
FIG. 1. The example demographic retriever 118 is structured to
retrieve the database proprietor demographic information 120a while
obscuring the identities of the members of the panelist households.
In the illustrated example, the demographic retriever 118 includes
an example decoy generator 200, an example panelist obscurer 202,
and an example query handler 204.
[0031] The example decoy generator 200 generates decoy DPIDs 122a,
122b used to reduce the likelihood that the database proprietor 110
is able to identify the panelist DPIDs 112 associated with the
members of the panelist households. The example decoy generator 200
creates PDFs to increase the likelihood that a decoy DPID 122a,
122b will correspond to a real subscriber to the database
proprietor 110 to better obscure the identities of members of the
panelist households. For example, if the database proprietor 110 is
trying to identify the members of the panelist households, the
database proprietor 110 would likely discard any queries associated
with decoy DPIDs 122b that do not correspond to real subscribers.
The example panelist DPIDs 112 have N digit positions. The example
decoy generator 200 creates PDFs that characterize the probability
that a current digit position (N.sub.j) has a particular value. The
number of digit positions (N) and a range of possible values (e.g.,
a value between 0 and 9 (decimal), a value between 0 and F
(hexadecimal), a value between 0 and Z (alphanumeric), etc.) may be
different for different database proprietors 110. For example, a
database proprietor 110 may have fourteen digit positions
(N.sub.0-N.sub.13) with possible decimal values (e.g., between 0
and 9). In such an example, a database proprietor 110 with 1.44
billion active subscribers has 100 trillion possible panelist DPIDs
112 that are potentially not assigned sequentially.
[0032] To generate the PDF for the corresponding digit positions
(N), the decoy generator 200, from time to time (e.g.,
periodically, aperiodically, etc.), statistically analyzes the
panelist DPIDs 112 in the DPID database 116. In some examples, the
decoy generator 200 also includes previously generated decoy DPIDs
122a that result in receiving database proprietor demographic
information 120b from the database proprietor 110 because those
ones of the decoy DPIDs 122a actually corresponded to real
subscribers. In the illustrated example, to generate a PDF (PDF)
for a digit position (N.sub.j), the decoy generator 200 calculates
independent probabilities (P.sub.ij) for the possible values in
that digit position (N.sub.j). For example, the decoy generator 200
may determine that the independent probability that the second
digit position is "7" is 14% (P.sub.i2(7)=14%)
[0033] In some examples, the decoy generator 200 calculates
conditional probabilities (P.sub.cj) for the possible values in
that digit position (N.sub.j) based on the value selected for a
previous digit position (N-1) and/or a next previous digit position
(N-2). For example, the decoy generator 200 may determine that the
condition probability that the second digit position is "7," given
that the value of the first digit position is "2," is 36%
(P.sub.c2(7|N.sub.1=2)=36%). In some examples, the decoy generator
200 calculates divergence between the independent probability
(P.sub.ij) and the conditional probability (P.sub.cj). The
divergence determines how dependent (e.g., conditional) the value
of the current digit position (N) on the value of a previous digit
position (e.g., N-1, N-2, etc.). In some such examples, the
divergence is calculated using the Jensen-Shannon divergence (JSD)
using Equation 1, Equation 2, and Equation 3 shown below.
JSD ( P cj P ij ) = 1 2 D ( P cj M ) + 1 2 D ( P ij M ) Equation 1
D ( P Q ) = i P ( i ) log 2 ( P ( i ) Q ( i ) ) Equation 2 M = 1 2
( P cj + P ij ) Equation 3 ##EQU00001##
In Equation 1, Equation 2, and Equation 3 above, JSD calculates the
extent that the value of the current digit position (N) is
conditional on the value of the previous digit position (N-1). In
the illustrated example, the decoy generator 200 compares the
calculated JSD to a divergence threshold. If the JSD satisfies
(e.g., is greater than) the divergence threshold, the decoy
generator 200 uses the conditional probability (P.sub.cj) based on
the previous digit position (N-1) when generating the corresponding
PDF (PDF.sub.j).
[0034] In some examples, if the JSD is not satisfied (e.g., is less
than) the divergence threshold, the decoy generator 200
recalculates the Jensen-Shannon divergence (JSD) using the
conditional probability (P.sub.cj) based on the other previous
digit position (e.g., N-2, N-3, etc.) until either (i) the
divergence threshold is satisfied, or (ii) the previous digit
positions have been tried. If a conditional probability (P.sub.cj)
based on one of the previous digit position satisfies the
divergence threshold, the particular conditional probability
(P.sub.cj) is used to generate the corresponding PDF (PDF.sub.j).
If the previous digit positions have been tried and the divergence
threshold has not been satisfied, the example decoy generator 200
uses the independent probability (P.sub.ij) to generate the
corresponding PDF (PDF.sub.j).
[0035] The example decoy generator 200 generates decoy DPIDs 122a,
122b by sampling the PDFs for the digit positions
(N.sub.0-N.sub.j). In some examples, the decoy generator 200 uses
inverse sampling to generate the decoy DPIDs 122a, 122b. For
example, for a fourteen-digit DPID, the decoy generator 200 samples
the fourteen PDFs corresponding to the fourteen digit positions.
The example decoy generator 200 assigns a confidence value to the
decoy DPIDs 122a, 122b based on the probability that a generated
decoy DPID 122a, 122b will correspond to a real subscriber to the
database proprietor 110. For example, if the decoy generator 200
randomly generates the decoy DPID 122a, 122b, the decoy generator
200 may assign the decoy DPID 122a, 122b a confidence level of
0.000014 (e.g., one out of every 70000 generated decoy DPIDs 122a,
122b is expected to correspond to a real subscriber registered with
the database proprietor 110). As another example, if the decoy
generator 200 generates the decoy DPID 122a, 122b with PDFs based
on independent probabilities (P.sub.IJ), the decoy generator 200
may assign the decoy DPID 122a, 122b a confidence level of 0.1
(e.g., one out of every ten generated decoy DPIDs 122a, 122b is
expected to correspond to a real subscriber registered with the
database proprietor 110).
[0036] In the illustrated example of FIG. 2, the panelist obscurer
202 receives or otherwise retrieves the decoy DPIDs 122a, 122b
generated by the decoy generator 200. The example panelist obscurer
202 randomly or pseudo-randomly distributes the panelist DPIDs 112
from the example DPID database 116 amongst the decoy DPIDs 122a,
122b to create the example batch query 123. FIG. 3 illustrated an
example batch query 123 with panelist DPIDs (e.g., the PDPIDs
112a-112d) randomly or pseudo-randomly distributed amongst the
decoy DPIDs (e.g., the DDPIDs 122a-122h). In the illustrated
example of FIG. 2, the AME 108 sets an obscuration target (O.sub.T)
that is the target ratio of decoy DPIDs 122a corresponding to real
subscribers and the panelist DPIDs 112. A higher obscuration target
(O.sub.T) makes it less likely that a database proprietor 110 would
be able to discern which of the queried DPIDs 112, 122a, 122b
belong to members of panelist households. For example, the
obscuration target (O.sub.T) may be 66%. In such an example, the
obscuration target (O.sub.T) of 66% means that for queries that
return database proprietor demographic information 120a, 120b, 34%
will correspond to the panelist DPIDs 112 and 66% will correspond
to decoy DPIDs 122a. A minimum number of the decoy DPIDs 122a, 122b
used by the panelist obscurer 202 to achieve the obscuration target
(O.sub.T) is calculated using Equation 4 below.
Min ( decoy DPID ) = ( Num ( Panelist DPID ) ( 1 - O T ) - Num (
Panelist DPID ) ) Confidence Level Equation 4 ##EQU00002##
In Equation 4 above, Min(decoy DPID) is the minimum number of the
decoy DPIDs 122a, 122b to be used, Num(Panelist DPID) is the number
of panelist DPIDs 112 to be obscured, and confidence level is the
confidence level assigned to the decoy DPIDs 122a, 122b by the
decoy generator 200. For example, if the panelist obscurer 202 is
to obscure 100 panelist DPIDs 112 with an obscuration target
(O.sub.T) of 66%, and the decoy generator 200 assigns the decoy
DPIDs 122a, 122b a confidence level of 0.1, the panelist obscurer
202 would use 1941 decoy DPIDs 122a, 122b
(((100/(1-0.66))-100)/0.1). In such an example, the panelist
obscurer 204 would randomly mix the 100 panelist DPIDs 112 into the
1941 decoy DPIDs 122a, 122b to create the batch query 123.
[0037] In the illustrated example of FIG. 2, the query handler 204
receives or otherwise retrieves the batch query 123 from the
example panelist obscurer 202. The example query handler 204
queries the database proprietor 110 with the randomly mixed
panelist DPIDS 112 and decoy DPIDs 122a, 122b in the example query
batch query 123 in the order presented in the batch query 123. In
the illustrated example, the query handler 204 uses an API provided
by the database proprietor 110 to query the database proprietor 110
for demographic information corresponding to the panelist DPIDs 112
and decoy DPIDs 122a, 122b. For the randomly mixed panelist DPIDs
112 and decoy DPIDs 122a, 122b, the example query handler 204 may
receive numerous types of responses from the database proprietor
110, including (i) database proprietor demographic information 120a
corresponding to the panelist PDIDs 112, (ii) database proprietor
demographic information 120b corresponding to the decoy DPIDs 122a,
and (iii) an error 125 corresponding to particular decoy DPIDs 122b
that are not associated with database proprietor demographic
information 120.
[0038] For panelist DPIDs 112 that correspond to database
proprietor demographic information 120a, the example query handler
204 forwards to the example panelist comparator 126 the panelist
DPIDs 112 that are associated with database proprietor demographic
information 120a along with the corresponding AME IDs 106 (e.g.,
the AME IDs 106 stored in association the panelist DPIDs 112 in the
DPID database 116). The example query handler 204 discards the
decoy DPIDs 122b that return with an error 125. In some examples,
the query handler 204 discards the decoy DPIDs 122a that are
associated with database proprietor demographic information 120b.
Alternatively, in some examples, the query handler 204 saves the
decoy DPIDs 122a that are associated with database proprietor
demographic information 120b to be used by the decoy generator 200
to, for example, update the PDFs used to generate the decoy DPIDs
122a, 122b. In this manner, the example decoy generator can improve
the PDFs by increasing the sample size of DPIDs 112, 122a used to
generate the PDFs. In such examples, the query handler 204 discards
the database proprietor demographic information 120b corresponding
to the decoy DPIDs 122a. In some examples, the decoy generator 200
uses the decoy DPIDs 122a that return the database proprietor
demographic information 120b and the decoy DPIDs 122b that return
the error 125 to adjust the confidence level used to calculate the
minimum number of decoy DPIDs 122a, 122b. In this manner, the
example panelist obscurer 202 can improve (e.g. decrease) the
number of decoy DPIDs 122a, 122b used to generate a batch query
123. For example, the decoy DPIDs 122a that return the database
proprietor demographic information 120b may indicate that the
confidence level is too low (e.g., one out of every fifteen decoy
DPIDs 122a correspond to an actual subscriber to the database
proprietor 110 instead of one out of every ten DPIDs 122a, 122b,
etc.).
[0039] While an example manner of implementing the example
demographic retriever 118 of FIG. 1 is illustrated in FIG. 2, one
or more of the elements, processes and/or devices illustrated in
FIG. 2 may be combined, divided, re-arranged, omitted, eliminated
and/or implemented in any other way. Further, the example decoy
generator 200, the example panelist obscurer 202, the example query
handler 204, and/or, more generally, the example demographic
retriever 118 of FIG. 1 may be implemented by hardware, software,
firmware and/or any combination of hardware, software and/or
firmware. Thus, for example, any of the example decoy generator
200, the example panelist obscurer 202, the example query handler
204, and/or, more generally, the example demographic retriever 118
could be implemented by one or more analog or digital circuit(s),
logic circuits, programmable processor(s), application specific
integrated circuit(s) (ASIC(s)), programmable logic device(s)
(PLD(s)) and/or field programmable logic device(s) (FPLD(s)). When
reading any of the apparatus or system claims of this patent to
cover a purely software and/or firmware implementation, at least
one of the example decoy generator 200, the example panelist
obscurer 202, and/or the example query handler 204 is/are hereby
expressly defined to include a tangible computer readable storage
device or storage disk such as a memory, a digital versatile disk
(DVD), a compact disk (CD), a Blu-ray disk, etc. storing the
software and/or firmware. Further still, the example the example
demographic retriever 118 of FIG. 1 may include one or more
elements, processes and/or devices in addition to, or instead of,
those illustrated in FIG. 2, and/or may include more than one of
any or all of the illustrated elements, processes and devices.
[0040] FIG. 4 illustrates the example panelist comparator 126 of
FIG. 1. The example panelist comparator 126 is structured to assign
a panelist DPID 112 to a member of a panelist household. In the
illustrated example, to make the assignment, the example panelist
comparator 126 uses the database proprietor demographic information
120a received from the demographic retriever 118 and panelist
demographic information 400 stored the panelist database 126. In
the illustrated example, the panelist comparator 126 includes an
example variant database 401, an example demographic comparator
402, and an example panelist associator 404. The example variant
database 401 includes variants of given names (e.g. nicknames, pet
names, diminutive forms, etc.). For example, for the given name
"Margaret," the variant database may include "Greta," "Maggie,"
Marge," "Margo," "Meagan," "Peg," "Peggy," and/or "Molly," etc.
[0041] The example demographic comparator 402 compares database
proprietor demographic information 120a with panelist demographic
information 400 in the panelist database 126. In the illustrated
example, the demographic comparator 402 receives the example AME ID
106, the example panelist DPID 112, and the example database
proprietor demographic information 120a from the demographic
retriever 118. In the illustrated example, the database proprietor
demographic information 120a includes a subscriber given name 406,
a subscriber date of birth (DOB) 408 and subscriber past name(s)
410. The example subscriber past name(s) 410 refer to names that
have been used in connection with the panelist DPID 112 in the past
according to the database proprietor demographic information 120a
(e.g., the database proprietor 110 of FIG. 1 stores past names when
a subscriber changes the name used in connection with the database
proprietor 110). In some examples, the database proprietor
demographic information 120a also includes other demographic
information (e.g., geographic location, race, income level,
education level, religion, etc.) that may be used by the
misattribution calculator 130 (FIG. 1) to calculate misattribution
correction factors.
[0042] The example demographic comparator 402 retrieves the
demographic information 400 of the members of the panelist
household identified by the AME ID 106. In the illustrated example,
the demographic information 400 includes example panelist given
names 412 and example panelist DOB s 414. The example demographic
comparator 402 standardizes the subscriber given name 406, the
subscriber past name(s) 410 and the panelist given name(s) 412 by
converting them to upper case characters and removing diacritics
(e.g. changing "a" to "a", changing "u" to "u", changing "n" to
"n", etc.).
[0043] The example demographic comparator 402 compares the
subscriber given name 406 to the panelist given name(s) 412 and/or
or variants of the panelist given name(s) 412 stored in the variant
database 400. In some examples, if the subscriber given name 406 is
not a match for the panelist given name(s) 412 and/or variants of
the panelist given name(s) 412, the example demographic comparator
402 compares the past name(s) 410 to the panelist given name(s) 412
and/or or variants of the panelist given name(s) 412. In some
examples, if neither the subscriber given name 406 nor the
subscriber past name(s) 410 are a match for one of the panelist
given names 412, the example demographic comparator 402 compares
the subscriber DOB 408 with the panelist DOB 414. In some examples,
if none of the subscriber given name 406, the subscriber past
name(s) 410, or the subscriber DOB 408 are a match, the example
demographic comparator 402 determines that the panelist DPID 112 is
not to be associated with the members of the panelist household
(e.g., the panelist DPID 112 may be associated with a friend of a
member of the panelist household, etc.). For example, if the
subscriber given name 406 is "Alex" and the subscriber past names
410 are "Joe" and "Joey," the demographic comparator 402 indicates
that the member of the panelist household with the panelist given
name 412 "Joseph" is a match. In that example, the demographic
comparator 402 indicates a match because variants (e.g., in the
variant database 401) of the panelist given name 412 "Joseph"
include "Joe" and "Joey," and "Joe" and "Joey" are the subscriber
past names 410 associated with the database proprietor demographic
information 120a.
[0044] In the illustrated example of FIG. 4, the panelist
associator 404 stores the panelist DPID 112 in the panelist
database 126 in association with the AME ID 106 of the member of
the panelist household that the example demographic comparator 402
indicated as a match for the panelist DPID 112. Additionally, the
example panelist associator 404 stores the database proprietor
demographic information 120a in association with the panelist DPID
112 in the panelist database 126.
[0045] While an example manner of implementing the example panelist
comparator 126 of FIG. 1 is illustrated in FIG. 4, one or more of
the elements, processes and/or devices illustrated in FIG. 4 may be
combined, divided, re-arranged, omitted, eliminated and/or
implemented in any other way. Further, the example demographic
comparator 402, the example panelist associator 404, and/or, more
generally, the example panelist comparator 126 of FIG. 1 may be
implemented by hardware, software, firmware and/or any combination
of hardware, software and/or firmware. Thus, for example, any of
the example demographic comparator 402, the example panelist
associator 404, and/or, more generally, the example panelist
comparator 126 could be implemented by one or more analog or
digital circuit(s), logic circuits, programmable processor(s),
application specific integrated circuit(s) (ASIC(s)), programmable
logic device(s) (PLD(s)) and/or field programmable logic device(s)
(FPLD(s)). When reading any of the apparatus or system claims of
this patent to cover a purely software and/or firmware
implementation, at least one of the example demographic comparator
402, and/or the example panelist associator 404 is/are hereby
expressly defined to include a tangible computer readable storage
device or storage disk such as a memory, a digital versatile disk
(DVD), a compact disk (CD), a Blu-ray disk, etc. storing the
software and/or firmware. Further still, the example the example
panelist comparator 126 of FIG. 1 may include one or more elements,
processes and/or devices in addition to, or instead of, those
illustrated in FIG. 2, and/or may include more than one of any or
all of the illustrated elements, processes and devices.
[0046] A flowchart representative of example machine readable
instructions for implementing the example demographic retriever 118
of FIGS. 1 and 2 is shown in FIG. 5. A flowchart representative of
example machine readable instructions for implementing the example
decoy generator 118 of FIG. 2 is shown in FIG. 6. A flowchart
representative of example machine readable instructions for
implementing the example panelist comparator of FIGS. 1 and 4 is
shown in FIG. 7. In these examples, the machine readable
instructions comprise a program for execution by a processor such
as the processor 812 shown in the example processor platform 800
discussed below in connection with FIG. 8. The program(s) may be
embodied in software stored on a tangible computer readable storage
medium such as a CD-ROM, a floppy disk, a hard drive, a digital
versatile disk (DVD), a Blu-ray disk, or a memory associated with
the processor 812, but the entire program and/or parts thereof
could alternatively be executed by a device other than the
processor 812 and/or embodied in firmware or dedicated hardware.
Further, although the example program is described with reference
to the flowchart illustrated in FIGS. 5, 6, and/or 7, many other
methods of implementing the example demographic retriever 118, the
example decoy generator 200, and/or the example panelist comparator
126 may alternatively be used. For example, the order of execution
of the blocks may be changed, and/or some of the blocks described
may be changed, eliminated, or combined.
[0047] As mentioned above, the example processes of FIGS. 5, 6,
and/or 7 may be implemented using coded instructions (e.g.,
computer and/or machine readable instructions) stored on a tangible
computer readable storage medium such as a hard disk drive, a flash
memory, a read-only memory (ROM), a compact disk (CD), a digital
versatile disk (DVD), a cache, a random-access memory (RAM) and/or
any other storage device or storage disk in which information is
stored for any duration (e.g., for extended time periods,
permanently, for brief instances, for temporarily buffering, and/or
for caching of the information). As used herein, the term tangible
computer readable storage medium is expressly defined to include
any type of computer readable storage device and/or storage disk
and to exclude propagating signals and to exclude transmission
media. As used herein, "tangible computer readable storage medium"
and "tangible machine readable storage medium" are used
interchangeably. Additionally or alternatively, the example
processes of FIGS. 5, 6, and/or 7 may be implemented using coded
instructions (e.g., computer and/or machine readable instructions)
stored on a non-transitory computer and/or machine readable medium
such as a hard disk drive, a flash memory, a read-only memory, a
compact disk, a digital versatile disk, a cache, a random-access
memory and/or any other storage device or storage disk in which
information is stored for any duration (e.g., for extended time
periods, permanently, for brief instances, for temporarily
buffering, and/or for caching of the information). As used herein,
the term non-transitory computer readable medium is expressly
defined to include any type of computer readable storage device
and/or storage disk and to exclude propagating signals and to
exclude transmission media. As used herein, when the phrase "at
least" is used as the transition term in a preamble of a claim, it
is open-ended in the same manner as the term "comprising" is open
ended.
[0048] FIG. 5 is a flow diagram representative of example machine
readable instructions 500 that may be executed to implement the
example demographic retriever 118 of FIGS. 1 and 2 and/or the
example panelist comparator 126 of FIGS. 1 and 4 to assign
demographic information to panelists. Initially, at block 502, the
example decoy generator 200 (FIG. 2) generates PDFs used to
generate the decoy DPIDs 122a, 122b (FIGS. 1, 2, and 3). An example
for generating the PDFs is disclosed in connection with FIG. 6
below. At block 504, the example decoy generator 200 generates the
decoy DPIDs 122. The example decoy generator 200 generates an
amount of the decoy DPIDs 122 as requested by the example panelist
obscurer 202 (FIG. 2). In some examples, the amount of decoy DPIDs
122a, 122b the panelist obscurer 202 requests is calculated in
accordance with Equation 4 above. At block 506, the example
panelist obscurer 202 randomly or pseudo-randomly distributes the
example panelist DPIDs 112 (FIGS. 1, 2, 3, and 4) retrieved from
computing devices 108 (FIG. 1) associated with panelist households
amongst the example decoy DPIDs 122a, 122b generated at block
504.
[0049] At block 508, the example query handler 204 (FIG. 2) queries
the database proprietor 110 (FIGS. 1 and 2) using the example
panelist DPIDs 112 randomly or pseudo-randomly distributed amongst
the example decoy DPIDs 122a, 122b. At block 510, the example query
handler 204 separates the results of the query between the example
panelist DPIDs 112 and the example decoy DPIDs 122a, 122b. At block
512, the example query handler 204 separates the example decoy
DPIDs 122a that correspond to returned database proprietor
demographic information 120b (FIGS. 1, 2, and 4) from example decoy
DPIDs 122b that correspond to the returned error 125 (FIGS. 1 and
2). In some examples, the example decoy DPIDs 122a that correspond
to returned database proprietor demographic information 120b are
saved for further analysis (e.g., to refine the PDFs generated by
the decoy generator 200, etc.).
[0050] At block 514, the example panelist obscurer 202 determines
whether there are more of the panelist DPIDs 112 retrieved from
example computing devices 108 associated with the panelist
households to be queried. If there are more of the panelist DPIDs
112, program control returns to block 504. Otherwise, if there are
not more of the panelist DPIDs 112, program control advances to
block 516. At block 516, the example panelist comparator 126
assigns the panelist DPIDs 112 to members of panelist households
stored in the panelist database 128 (FIGS. 1 and 4). An example of
assigning the panelist DPIDs 112 to members of panelist households
is disclosed in connection with FIG. 7 below. The example program
500 then ends.
[0051] FIG. 6 is a flow diagram representative of example machine
readable instructions 502 that may be executed to implement the
example decoy generator 200 of FIG. 2 to generate PDFs used to
generate decoy DPIDs 122a, 122b (FIGS. 1 and 2). Initially, at
block 602, the example decoy generator 200 selects a digit position
(N). For example, for a database proprietor (e.g., the database
proprietor 110 of FIGS. 1 and 2) that has a ten digit DPID (e.g.,
digits N.sub.0-N.sub.9), the decoy generator 200 initially selects
the digit position with the smallest place value (e.g., digit
N.sub.0) of a decoy DPID 122a, 122b to be generated. At block 604,
the decoy generator 200 calculates independent probabilities (P(N))
for the possible values (e.g., 0-9, etc.) of the current digit
position (N). For example, the decoy generator 200 may calculate a
number of times a value appears in the current digit position (N)
divided by a total number of DPIDs being analyzed. For example, the
current digit (N.sub.0) may have independent probabilities as
illustrated in Table 1 below.
TABLE-US-00001 TABLE 1 EXAMPLE INDEPENDENT PROBABILITIES FOR THE
CURRENT DIGIT POSITION N.sub.0 Value 0 1 2 3 4 5 6 7 8 9 P(N.sub.0)
0.10 0.05 0.05 0.30 0.15 0.12 0.03 0.00 0.15 0.05
[0052] At block 606, the decoy generator 200 calculates first
conditional probabilities (P(V|N-1)) of the current digit position
(N) based on the value of the previous digit position (N-1). For
example, the decoy generator 200 may calculate a number of times a
value appears in the current digit position (N) when the previous
digit position (N-1) has a particular value divided by the total
number of DPIDs being analyzed. For example, the current digit
(N.sub.1) may have conditional probabilities based on the previous
digit (N.sub.0) as illustrated in Table 2 below.
TABLE-US-00002 TABLE 2 EXAMPLE CONDITIONAL PROBABILITIES FOR THE
CURRENT DIGIT POSITION (N.sub.1) BASED ON PREVIOUS DIGIT (N.sub.0)
Value (V) 0 1 2 3 4 5 6 7 8 9 P(V|N.sub.0 = 0) 0.10 0.05 0.05 0.30
0.15 0.12 0.03 0.00 0.15 0.05 P(V|N.sub.0 = 1) 0.08 0.03 0.04 0.25
0.27 0.18 0.00 0.00 0.12 0.03 . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . P(V|N.sub.0 = 9) 0.11 0.05 0.05 0.31
0.14 0.12 0.00 0.00 0.15 0.07
[0053] At block 608, the decoy generator 200 calculates the
divergence between the independent probabilities generated at block
604 and the first conditional probabilities generated at block 606.
The example decoy generator 200 calculates the Jensen Shannon
divergence in accordance with Equation 1, Equation 2, and Equation
3 above. At block 610, the example decoy generator 200 determines
if the divergence calculated at block 608 satisfies (e.g., is
greater than) a divergence threshold. If the divergence satisfies
the divergence threshold, program control advances to block 612.
Otherwise, if the divergence does not satisfy the divergence
threshold, program control advances to block 614. At block 612, the
example decoy generator 200 generates the PDFs for the current
digit position (N) based on the first conditional probabilities
calculated at block 606. Program control then advances to block
624.
[0054] At block 614, the decoy generator 200 calculates second
conditional probabilities (P(V|N-2)) of the current digit position
(N) based on the value of the next previous digit position (N-2).
For example, the decoy generator 200 may calculate a number of
times a value appears in the current digit position (N) when the
next previous digit position (N-1) has a particular value divided
by the total number of DPIDs being analyzed. At block 616, the
decoy generator 200 calculates the divergence between the
independent probabilities generated at block 604 and the second
conditional probabilities generated at block 614. At block 618, the
example decoy generator 200 determines it the divergence calculated
at block 616 satisfies (e.g., is greater than) the divergence
threshold. If the divergence satisfies the divergence threshold,
program control advances to block 620. Otherwise, if the divergence
does not satisfy the divergence threshold, program control advances
to block 622. At block 620, the example decoy generator 200
generates the PDFs for the current digit position (N) based on the
second conditional probabilities calculated at block 614. Program
control then advances to block 624.
[0055] At block 622, the example decoy generator 200 generates the
PDFs for the current digit position (N) based on the independent
probabilities calculated at block 604. At block 624, the example
decoy generator 200 determines whether there is another digit
position (N+1) to analyze. If there is another digit position to
analyze, program control returns to block 602 to select the next
digit position of the decoy DPID being generated. Otherwise, if
there is not another digit position to analyze, the example program
502 ends.
[0056] FIG. 7 is a flow diagram representative of example machine
readable instructions 516 that may be executed to implement the
example panelist comparator 126 of FIGS. 1 and 4 to assign panelist
DPIDS 112 (FIGS. 1, 2, 3, and 4) to members of a panelist
household. Initially, at block 702, the example demographic
comparator 402 selects one of the panelist DPIDs 112 received or
otherwise retrieved from the example demographic retriever 118. At
block 704, the example demographic comparator 402 retrieves
panelist demographic information 400 (FIG. 4) (e.g., the panelist
given name(s) 412 (FIG. 4), the panelist DOB(s) 414 (FIG. 4), etc.)
of member(s) of the panelist household identified by the AME ID 106
(FIGS. 1 and 4) associated with the selected panelist DPID 112 from
the example panelist database 126 (FIGS. 1 and 4).
[0057] At block 706, the example demographic comparator 402
standardizes given names. For example, the demographic comparator
402 standardizes the example subscriber given names 406 (FIG. 4)
and the example subscriber past names 410 (FIG. 4) included in
example database proprietor demographic information 120a associated
with the example panelist DPID 112 and the example panelist given
name(s) 412 associated with the example AME ID 106. To standardize
the example subscriber given names 406, the example subscriber past
names 410 and the example panelist given name(s) 412, the example
demographic comparator 402 capitalizes letters and removes the
diacritics. For example, the subscriber given name 406 "Jozsef"
would be standardized to "JOZSEF."
[0058] At block 708, the example demographic comparator 402
determines whether the subscriber given name 406 matches one of
more of the panelist given names 412 and variants of the panelist
given name 412 stored in the variant database 401. If the
subscriber given name 406 matches one of the panelist given names
412 or one of the variants of the one of panelist given names 412,
program control advances to block 716. Otherwise, if the subscriber
given name 406 does not match the panelist given names 412 or the
variants of the panelist given names 412, program control advances
to block 710. At block 710, the example demographic comparator 402
compares the subscriber past name(s) 410 to the panelist given
name(s) 412 and the variants of the panelist given names 412. If
one of the subscriber past names 410 matches one of the panelist
given names 412 or one of the variants of one of the panelist given
names 412, program control advances to block 716. Otherwise, if the
subscriber past names 410 do not match the panelist given names 412
or the variants of the panelist given names 412, program control
advances to block 712. At block 712, the example demographic
comparator 402 compares the subscriber DOB 408 with the panelist
DOB(s) 414. If the subscriber DOB 408 matches one of the panelist
DOBs 414, program control advances to block 716. If the subscriber
DOB 408 does not match the panelist DOBs 414, program control
advances to block 714. At block 714, the example panelist
associator 404 determines that the panelist DPID 112 selected at
block 702 does not belong to a member of the panelist
household.
[0059] At block 716, the example panelist associator 404 associates
the panelist DPID 112 with the matching member of the panelist
household determined at block 708, block 710, or block 712. The
example panelist associator 404 also associates the database
proprietor demographic information 120a with the matching member of
the panelist household. Additionally, the example panelist
associator 404 stores the panelist DPID 112 and the database
proprietor demographic information 120a in the panelist database
128 in association with the AME ID 106. At block 718, the example
demographic comparator 402 determines whether another panelist DPID
112 is to be compared to the members of the panelist households in
the panelist database 128. If another panelist DPID 112 is to be
compared, program control returns to block 702. Otherwise, if
another panelist DPID 112 is not to be compared, the example
program 516 ends.
[0060] FIG. 8 is a block diagram of an example processor platform
800 is structured to execute the instructions of FIGS. 5, 6, and/or
7 to implement the example demographic retriever 118 of FIGS. 1 and
2, the example decoy generator 200 of FIG. 2, and/or the example
panelist comparator 126 of FIGS. 1 and 4. The processor platform
800 can be, for example, a server, a personal computer, a
workstation, or any other type of computing device.
[0061] The processor platform 800 of the illustrated example
includes a processor 812. The processor 812 of the illustrated
example is hardware. For example, the processor 812 can be
implemented by one or more integrated circuits, logic circuits,
microprocessors or controllers from any desired family or
manufacturer. In the illustrated example, the processor 812
includes the example demographic retriever 118 with the example
decoy generator 200, the example panelist obscurer 202, and the
example query handler 204. The example processor 812 of the
illustrated example also includes the example panelist comparator
126 with the example demographic comparator 402 and the example
panelist associator 404.
[0062] The processor 812 of the illustrated example includes a
local memory 813 (e.g., a cache). The processor 812 of the
illustrated example is in communication with a main memory
including a volatile memory 814 and a non-volatile memory 816 via a
bus 818. The volatile memory 814 may be implemented by Synchronous
Dynamic Random Access Memory (SDRAM), Dynamic Random Access Memory
(DRAM), RAMBUS Dynamic Random Access Memory (RDRAM) and/or any
other type of random access memory device. The non-volatile memory
816 may be implemented by flash memory and/or any other desired
type of memory device. Access to the main memory 814, 816 is
controlled by a memory controller.
[0063] The processor platform 800 of the illustrated example also
includes an interface circuit 820. The interface circuit 820 may be
implemented by any type of interface standard, such as an Ethernet
interface, a universal serial bus (USB), and/or a PCI express
interface.
[0064] In the illustrated example, one or more input devices 822
are connected to the interface circuit 820. The input device(s) 822
permit(s) a user to enter data and commands into the processor 812.
The input device(s) can be implemented by, for example, an audio
sensor, a microphone, a camera (still or video), a keyboard, a
button, a mouse, a touchscreen, a track-pad, a trackball, isopoint
and/or a voice recognition system.
[0065] One or more output devices 824 are also connected to the
interface circuit 820 of the illustrated example. The output
devices 824 can be implemented, for example, by display devices
(e.g., a light emitting diode (LED), an organic light emitting
diode (OLED), a liquid crystal display, a cathode ray tube display
(CRT), a touchscreen, a tactile output device, a printer and/or
speakers). The interface circuit 820 of the illustrated example,
thus, typically includes a graphics driver card, a graphics driver
chip or a graphics driver processor.
[0066] The interface circuit 820 of the illustrated example also
includes a communication device such as a transmitter, a receiver,
a transceiver, a modem and/or network interface card to facilitate
exchange of data with external machines (e.g., computing devices of
any kind) via a network 826 (e.g., an Ethernet connection, a
digital subscriber line (DSL), a telephone line, coaxial cable, a
cellular telephone system, etc.).
[0067] The processor platform 800 of the illustrated example also
includes one or more mass storage devices 828 for storing software
and/or data. Examples of such mass storage devices 828 include
floppy disk drives, hard drive disks, compact disk drives, Blu-ray
disk drives, RAID systems, and digital versatile disk (DVD)
drives.
[0068] Coded instructions 832 of FIGS. 5, 6, and/or 7 may be stored
in the mass storage device 828, in the volatile memory 814, in the
non-volatile memory 816, and/or on a removable tangible computer
readable storage medium such as a CD or DVD.
[0069] From the foregoing, it will be appreciated that examples
have been disclosed which allow the AME to retrieve demographic
information of members of panelist households from database
proprietors while protecting the privacy of the members of panelist
households. In some examples, computers operate more efficiently by
generating a relatively small number of decoy DPIDs using the
disclosed PDFs compared to selecting decoy DPIDs at random. For
example, to obscure 200 panelist DPIDs with a 50% obscuration
target, by randomly selecting decoy DPIDs, the computer would need
to generate 14 million decoy DPIDs. In such an example, using the
methods, apparatus, and/or articles of manufacture disclosed
herein, the computer would need to generate 2000 decoy DPIDs.
Additionally, in such examples, by querying the database proprietor
using fewer decoy DPIDs, the methods, apparatus, and/or articles of
manufacture reduce bandwidth usage. Additionally, it will be
appreciated that examples have been disclosed which allow the AME
to assign demographic information retrieved from the database
proprietors to members of panelist households.
[0070] Although certain example methods, apparatus and articles of
manufacture have been disclosed herein, the scope of coverage of
this patent is not limited thereto. On the contrary, this patent
covers all methods, apparatus and articles of manufacture fairly
falling within the scope of the claims of this patent.
* * * * *